【摘要】 天然常绿阔叶混交林是宁波地区的典型地带性植被类型,研究该林分类型的树高-胸径关系模型,可为宁波地区的天然常绿阔叶混交林的林木生长预测、森林经营管理、森林资源清查等提供理论参考。以宁波地区常见的石栎 Lithocarpus glaber-木荷Schima superba天然常绿阔叶混交林为研究对象,选用30个常用的树高-胸径经验模型,通过5个模型评价指标即调整决定系数(Ra~2)、均方根误差(RMSE)、平均绝对误差(MAE)、相对平均绝对误差(RMAE)和Akaike信息准则(AIC)比较分析模型之间的拟合效果差异,从而确定适宜的树高-胸径模型。结果表明,除M23无法求解,其余29个模型均可求解。29个模型中,M09、M12和M26在5个模型评价指标中表现优异,M09和M12优于M26,说明M09和M12均为适宜的石栎-木荷天然常绿阔叶混交林树高-胸径模型。若优先考虑建模数据的拟合效果,建议选用M09,Ra~2=0.914 1;若优先考虑检验数据的拟合效果,建议选用M12,Ra~2=0.912 6。更多还原

【Abstract】 Natural evergreen broad-leaved mixed forests are typical zonal vegetation types in Ningbo.Height-diameter models for natural evergreen broad-leaved mixed forests may provide a theoretical basis for forest growth prediction,forest management,forest inventory,et al.30 height-diameter empirical models were constructed for Lithocarpus glaber-Schima superba natural evergreen broad-leaved mixed forest in Ningbo.The differences of fitting effect were analyzed among 30 height-diameter empirical models to determinate the appropriate height-diameter models using five model evaluation indices,namely,adjusted coefficient of determination (Ra~2),root mean square error (RMSE),mean absolute error (MAE),relative mean absolute error (RMAE) and Akaike information criterion (AIC).The results show that 29 models can be solved except for M23.The fitting effects of M09,M12 and M26 are better among 29 models using five model evaluation indices.Additionally,M09 and M12 are better than M26.M09 and M12 are the optimization models for L.glaber-S.superba natural evergreen broad-leaved mixed forest.Priority of calibration fitting effects,M09 is recommended,Ra~2=0.9141;Priority of validation fitting effects,M12 is recommended,Ra~2=0.9126.更多还原